J. Cent. South Univ. (2021) 28: 2193-2205

DOI: https://doi.org/10.1007/s11771-021-4762-2

Experimental study on bond behavior between BFRP bars and seawater sea-sand concrete

SU Xun(苏迅)^{1, 2}, YIN Shi-ping(尹世平)^{1, 2}, ZHAO Ying-de(赵瑛德)², HUA Yun-tao(华云涛)²

1. Jiangsu Key Laboratory of Environmental Impact and Structural Safety in Engineering, School of Mechanics & Civil Engineering, China University of Mining and Technology, Xuzhou 221116, China;

2. State Key Laboratory for Geomechanics & Deep Underground Engineering, School of Mechanics &Civil Engineering, China University of Mining and Technology, Xuzhou 221116, China

Central South University Press and Springer-Verlag GmbH Germany, part of Springer Nature 2021

Abstract: Combining fiber reinforced polymer (FRP) with seawater sea-sand concrete (SSC) can solve the shortage of river sand that will be used for marine engineering construction. The bond performance of BFRP bars and SSC specimens is researched by pull-out test in this paper. The effects of the parameters, such as bar type, bar diameter, concrete type and stirrup restraint, are considered. It is beneficial to the bonding performance by the reduction of bar diameter. The utilization of seawater sea-sand has a low influence on the bond properties of concrete. The bond strength of BFRP is slightly lower than the steel rebar, but the difference is relatively small. The failure mode of the specimen can be changed and the interfacial bond stress can be improved by stirrups restraint. The bond-slip curves of BFRP ribbed rebar include micro slip stage, slip stage, descent stage and residual stage. The bond stress shows the cycle attenuation pattern of sine in the residual stage. In addition, the bond-slip model of BFRP and SSC is obtained according to the experimental results and related literature, while the predicted curve is also consistent well with the measured curve.

Key words: basalt fiber-reinforced polymer (BFRP); seawater sea-sand concrete (SSC); bond-slip curve; constitutive model

Cite this article as: SU Xun, YIN Shi-ping, ZHAO Ying-de, HUA Yun-tao. Experimental study on bond behavior between BFRP bars and seawater sea-sand concrete [J]. Journal of Central South University, 2021, 28(7): 2193-2205. DOI: https://doi.org/10.1007/s11771-021-4762-2.

1 Introduction

In recent years, with the development of urbanization in coastal areas and the development of oceanic islands and reefs, there is a lack of river sand in some areas. In addition, the transportation of raw materials over long distances not only increases transportation costs, but impacts the process of the project. In view of the world’s vast seas and abundant sea-sand reserves, the utilization of seawater sea-sand has little influence on the mechanical performance of concrete, so the seawater sea-sand concrete (SSC) can be used in engineering construction [1, 2]. Fiber reinforced polymer (FRP) bar is a non-metallic reinforced material consisting of fiber and resin, and it has excellent chloride corrosion resistance [3, 4]. Thus, in the construction of marine engineering, the SSC structure reinforced by FRP bars can avoid corrosion of steel rebars and significantly enhance the durability of the structure.

The bond performance is the basis to ensure the effective work of FRP-SSC members, which affects not only the loading capacity and failure mode of the structure, but also the service performance such as cracks and deflections. At the moment, scholars have conducted many studies about the bond behavior of FRP bars and ordinary concrete [5-12]. Generally,FRP bars are made of anisotropic materials. The transverse performance of FRP bars is controlled by resins, and its longitudinal performance is controlled by fibers [13]. This feature causes that the bond property between the concrete and FRP bar has certain difference, with different FRP bars [7, 8] indicated that the bond strength of BFRP declines with the increase of BFRP bar diameter and bond length. The surface shape of the bars has a partial influence on the interfacial frictional and mechanical bite force. ROLLAND et al [9] noted that sand-coated BFRP bars can enhance effectually the initial stiffness and bonding strength of the interface. However, GU et al [10] noted that rib depth has a significant influence on bond slip, and it is considered that glass fiber-reinforced polymer (GFRP) bars with rib depth of 1 mm and rib spacing of 0.8d (diameter) can provide better bonding performance. In addition, the cover thickness also affects the form of bond failure. MAZAHERIPOUR et al [11] showed that the shear resistance on the surface of GFRP bars controlled the bond failure, and a lager cover thickness provided better restraint and higher bond strength. Under corrosive environments, ALTALMAS et al [12] pointed out that the initial bonding force increased because of the increase in concrete strength and the expansion of BFRP bars caused by corrosion; however, the final bonding strength reduced and the slip was basically unchanged. In addition, the bond-slip model can intuitively reflect the interfacial properties. At present, typical bond-slip models mainly include BPE model [14], MBPE model [15], CMR model [16], continuous curve model [17] and GFRP/steel bars strand composite bar bond-slip model [18], etc. Most existing models are based on GFRP bars and CFRP bars, and to some extent, the models do not accurately reflect the complete stress process and test curve characteristics of FRP bars under pull-out load.

Nowadays, the integration of FRP bars and seawater sea-sand is of widespread concern [19-23][20][21][22]. As mentioned above, the bond between FRP bars and SSC is crucial to the structure. XIAO et al [24] pointed out that the inclusion of recycled coarse aggregates reduced the bonding property of sea-sand regenerated concrete with GFRP bars, but the utilization of seawater and sea-sand had little impact on the bonding performance. XU et al [25] also drew similar conclusions. In terms of durability, Dong et al[26] found that in chloride environment, the bond failure mode of BFRP bar changed from interlaminar shear failure to splitting failure of bar surface, and the interfacial deterioration in seawater immersion conditions was larger than that in wet-dry cycle. DONG et al [27] further studied by eccentric pull-out test method and noted that the bond performance between SFCB bars and SSC decreases with the prolongation of erosion time, however,this is by design.

Summing up, the interfacial research of FRP bar SSC has just begun. Some research factors such as restraint of stirrups have not been involved and systematic conclusions have not been formed. In addition, the bond-slip model of BFRP-SSC needs to be further determined. To facilitate the advancement of BFRP-SSC structure, the influence of bar type, bar diameter, stirrup restraint and concrete type on bonding property are considered in this paper. And the bond-slip model of ribbed BFRP bars with SSC has been obtained and its applicability has been verified.

2 Experimental projects

2.1 Material characteristics

BFRP bars used in the test were supplied by Jiangsu Green Materials Vally New Material T&D Co., Ltd (GMV). Continuous basalt fibers and vinyl resins are wetted and cured through the extrusion process. The diameters include 8, 10, 12 and 16 mm. In addition, a 12 mm diameter ribbed bar was used as the control group. Ribbed BFRP is an integral deep rib with a rib spacing equal to the diameter and a rib depth is 0.06 times the diameter. The BFRP stirrups have a diameter of 8 mm and a length of 102 mm×102 mm at the center edge. The bar material is shown in Figure 1. According to Ref. [28], the tensile properties of FRP bars were tested, and the basic mechanical performance indexes of BFRP bars were shown in Table 1.

Figure 1 FRP bars used in test (Unit: mm):

Table 1 Mechanical properties of bars

The concrete matrix of the pull-out specimens consists of SSC and ordinary concrete. The compressive strength of concrete is designed to be 40 MPa. Artificial seawater is prepared according to Ref. [29], and in Table 2, the chemical composition details are given. The sea-sand from Qingdao, Shandong Province, was purchased with continuous medium sand grade and 0.06% chloride ion content. The concrete mix ratio is shown in Table 3. For SSC, the average compressive strength of cubic specimens (100 mm×100 mm×100 mm) after 28 days of curing is 39.34 MPa, the axial compressive strength of prismatic specimens (150 mm×150 mm×300 mm) is 31.08 MPa and the elastic modulus is 32.43 GPa. Meanwhile, for ordinary concrete made with general batched water along with the river sand, the 28 d compressive strength of cubic specimens is 40.36 MPa, the axial compressive strength is 31.88 MPa and the elastic modulus is 32.68 GPa. The 28 d compressive strength of normal concrete is somewhat higher than that of SSC, which is consistent with the conclusions in literature [1, 2].

2.2 Pull-out specimen

Generally, pull-out tests are used to evaluate bonding properties due to their easy operation and easy measurement of the free end slip. Refer to Ref. [30] and [31], the center pull-out test method is used in this test. The specimen size was 150 mm× 150 mm×150 mm. The factors considered include diameter (8, 10, 12 and 16 mm), type of bars (BFRP bars and rebars), type of concrete (SSC and ordinary concrete), and constraint of stirrups (number). Therefore, each group of 5, a total of 10 groups of 50 cubic pull-out specimens were made. The specimens are named “ABC-D-E”, where “A” represents the type of concrete, “B” represents the bar type, “C” represents the bar diameter, “D” represents the bond length, and “E” represents the amounts of stirrups. For example, “SWB12-4d-R2” refers to a seawater sea-sand concrete specimen consisting of 12 mm diameter BFRP bars with a 4d (diameter) bond length and a number of stirrups of 2.

The length of FRP bar is 550 mm and the free end is 50 mm. In order to avoid shear failure at the loading side, the loading side is anchored with 250 mm stainless steel pipe filled with epoxy resin and hardener. In order to control the bond length, PVC plastic sleeves were set at the free side of BFRP bars and the loading side respectively. The PVC plastic sleeves can not only avoid concrete crushing damage at the load side, but enable a more uniform distribution of bond stress. For specimens restrained by BFRP stirrups, when there is one stirrup, it is placed in the middle of the bonding section; when there are two stirrups, they are placed respectively at both end of the bonding section. The drawing of the pull-out specimen is shown in Figure 2.

The pull-out test was carried out on a 1000 kN electro-hydraulic servo universal testing machine loaded, and the device for loading is shown in Figure 3. With displacement control, the rate of loading is 0.5 mm/min. At the upper end of the cage device, a load sensor was placed to measure the tensile force, and displacement gauges were placed on the free and loaded ends of the specimen to measure the relative slip of BFRP bars and SSC.

Bond performance may be directly assessed by the bond strength. In general, the bond force distribution is not uniform along the bond length, but due to the short bond length in this test (equivalent to 5 times the bar diameter), the bond stress can be considered to be uniformly distributed during the test. The average bond stress is therefore determined as [8, 9, 24-26]:

                                 (1)

where P is the pull-out load (kN); D is the bar diameter (mm); l is the bonding length (mm).

Table 2 Chemical composition of artificial seawater

Table 3 Concrete mix proportion (kg/m³)

Figure 2 Drawing specimen diagram (Unit: mm)

Figure 3 Pull-out test loading device

3 Test results

The peak bond strengths, corresponding free end slips of the pull-out specimens and the failure modes of each specimen are shown in Table 4.

3.1 Failure modes

The pull-out specimen failure modes include splitting failure, splitting-pull failure and pull-out failure, as shown in Table 4 and Figure 4. Most specimens are pull-out failure, which is an ideal failure mode with certain ductility characteristics. For pull-out damage, the BFRP bars slip in the SSC, while the SSC test block remains intact, and the bars are finally pulled out of the SSC (Figure 4(a)). As shown in Figures 5(a) and (b), because of the low shear strength of the cross ribs of BFRP bars, the ribs are worn and the surface fibers and resin are shed off more. Related literatures have similarly reported [7, 11, 32], showing that the bond strength depends on the interlaminar shear strength of the inner core and outer part of FRP bar. However, the hardness of the steel rebar and shear strength of transverse ribs are higher, concrete between ribs is sheared out, and obvious corrosion can be seen on the surface of reinforcement. The splitting failure was largely focused on the 16 mm diameter bonded specimens. The slip between bars and concrete is small, and that is in essence the splitting failure of the concrete. Larger diameter of bars results in greater mechanical biting force. When the circumferential tensile stress formed by extrusion of the cross ribs of BFRP bars to the surrounding concrete exceeds the ultimate tensile strength of SSC, cracks occur in the concrete [33]. The failure is more sudden, cracking noise is emitted, and the bar is completely separated from the concrete base. The specimens were split into 2 pieces centered on the FRP bar (Figure 4(b)). At the location of the bonding section, clear FRP rib marks remain on the concrete matrix and the interfacial coarse aggregate occurred fracture. There is no abrasion on the surface of FRP bar. A small amount of SSC powder is adhered to the cross ribs (Figure 5(c)). Splitting-pullout damage is mainly concentrated on 16 mm diameter BFRP bar specimens restrained by stirrups. The development of cracks in SSC is limited by the reinforcement, which makes the failure mode change from splitting failure to splitting-pulling failure. The failure theory is analogous to splitting failure, however, the essence is bonding anchorage failure. The specimen surface shows fine cracks without obvious splitting. FRP bar and SSC are still integral and not fully separated (Figure 4(c)). It resembles pull-off failure, where the fibers on the surface of the bar partially fall off and the resin fiber powder remains in concrete.

Table 4 Pull-out test results

Continued

Figure 4 Failure modes of pull-out specimens:

Figure 5 Bars surface after specimen destruction:

3.2 Bond-slip curves

The bond-slip curve can more intuitively reflect the change of bond stress of FRP bar during pull-out. In general, the slip at the free side of the specimen always lags behind the slip of the loaded side during loading. Considering the large force of BFRP bar at loading end, obvious deformation occurs and there may be contact deviation between pull-out specimen and cage loading device. It is therefore reasonable to select the displacement of the free end of the specimen to reflect the slip [8, 9, 24, 26, 27, 32]. The most representative bond-slip curves for each group are chosen to analyze for different study factors as shown in Figure 6.

The complete stress process of ribbed BFRP bars and SSC can be simplified into micro slip stage, slip stage, descent stage and residual stage when pull-out failure occurs. The bonding stress during the micro slip stage is primarily supplied by the chemical bond between the two. But chemical adhesives usually exert little effort and the duration of this process is short or almost none [8, 33].

Figure 6(a) illustrates the bond-slip curves of different diameters specimens. At the beginning of loading, during the micro-slip stage, the curve slope of each specimen is close. The bond-slip curve of BFRP bars with smaller diameter has a longer micro slip stage and a higher slope of the curve during the slip stage. When the diameter of reinforcing bar is large, the slope at the slip stage tends to decrease gradually and the amount of slip increases continuously. This is mainly because the rib spacing of the smaller diameter FRP bars are narrow, the absolute value of stress/slip is larger, which leads a sharp curve. In addition, for the descent stage, the slip curve of the specimens with smaller diameter of bar decreases faster, which indicates that the damage of small diameter BFRP rib teeth is more serious at higher bond strength.

Figure 6(b) shows the bond-slip curves of different concrete types. The upper section of bond-slip curve of BFRP bar ordinary concrete pull-out specimens coincides with that of BFRP bar SSC specimens, shown that the utilization of seawater sea-sand has no remarkable impact on the bond-slip relationship of BFRP bars with concrete, which is consistent with the conclusion of reference [24, 25]. Hence, the same constitutive model as ordinary concrete can be used for the micro slip stage and the slip stage of bond-slip curve of BFRP bars SSC. Moreover, it was also confirmed that a small amount of shell in seawater sea-sand has no significant influence on the bonding behavior in this experiment.

Figure 6(c) shows the bond-slip curves of specimens of each type of bars. Overall, there is no residual stage in the bond-slip curve of steel rebar specimens. However, the bond slip-curve of the BFRP bar specimens has a residual stage and the bond stress decreases cyclically. BFRP bars are very close to the slope of steel rebars in the micro slip stage, which indicates that there is no apparent distinction in chemical adhesive force. The slope of the curve in the slip stage of the steel rebar specimens is a little greater than BFRP bar specimens, which could be associated with the elastic modulus of the steel rebar and the surface morphology of steel bar [33]. Since the elastic modulus of the steel rebar is considerably larger than BFRP bar, a larger bond can be obtained with a smaller amount of deformation. Besides, the gradient of transverse rib of BFRP bar changes more smoothly than that of steel rebar, and the bar material needs more slip amount in SSC to get the same mechanical interaction force as the steel rebar. During the descending stage, due to the obvious transverse rib biting action of reinforcement, the occlusal wear of SSC is more serious than that of BFRP bar, and the mechanical biting force at the interface loses more, thus the curve decreases faster.

Figure 6 Bond-slip curves of different specimens:

Figures 6(d) and (e) show the bond-slip curves of the stirrups constrained specimens. In the micro slip stage and the early stage of the slip, the bond-slip curves of each specimen are basically consistent. For stirrup restrained specimens, the slip stage is longer in the later stage with correspondingly higher peak bond stress. The slip of the specimen with stirrups constraint is marginally smaller than the slip of the specimen without the stirrup restraint. During the descend stage, the curve of the stirrups constrained specimen is gentler, and the bond stress in the residual stage is also increased accordingly. For specimens with 16 mm diameter of BFRP bar, as the number of FRP bars increases, the failure pattern of specimens changes to splitting-pulling failure from splitting failure, which is further converted to pull-out failure. This is primarily due to the strong stirrup constraint on the core SSC of the pull-out specimens, which improves the bonding performance [26, 27].

3.3 Bond strength analysis

3.3.1 Effect of bar diameter

From Table 4 and Figure 6(a), it is observed that the pull-out load of the specimens progressively increases as the diameter of the BFRP bar increases, but the bond strength keeps decreasing continuously. Compared with 8 mm diameter BFRP bar specimens, the bond strength of 10, 12 and 16 mm diameter BFRP bar specimens decreased by 15.8%, 29.4% and 56.7%, respectively.

The above phenomena are mainly attributed to Poisson effect, the shear lag phenomenon and bleeding of seawater sea-sand concrete [8, 9]. First of all, BFRP bars are anisotropic materials, the tensile strength is determined by the longitudinal fibers and the transverse shear strength is determined by the resins. Substantial elongation of BFRP bars in the bonding section will result in a decrease in interfacial friction (radial interfacial pressure of the bars decreases as the pull-out load increases) and a decline in bond strength. Secondly, the stresses and strains in the section of the BFRP bars are not uniformly distributed under tension. Generally, the deformation at the edge of bar is much larger than that at the center of the section of bar. The surficial bond stress of bar actually reflects the bond efficiency between bar and concrete. However, the average bond strength used to measure bonding properties is inherently small, and as the diameter increases, the difference between the two becomes more and more obvious. Finally, when pouring the specimens, the concrete may produce voids on the lower surface of the bar material because of bleeding, and the larger the diameter of the BFRP bar will aggravate the occurrence of bleeding phenomenon, the larger the void formed. This reduces the contact area between BFRP bars and SSC and weakens the bonding performance of both.

3.3.2 Effect of concrete type

As seen in Table 4 and Figure 6(b), the bond strength of BFRP bars and SSC is comparable to BFRP bars and ordinary concrete, which is 23.54 and 22.58 MPa respectively. Visibly, the influence of replacing freshwater river sand with seawater sea-sand on the bond performance of the BFRP bars to concrete can be negligible in the short term.

The main reason for the above phenomena is that the basic mechanical performance of SSC is comparable to ordinary concrete. The related studies show that the SSC has a higher early compressive strength than ordinary concrete, while the later compressive strength develops slower than that of ordinary concrete, but there is no remarkable distinction between the 28 d compressive strength of the two [1, 2]. In addition, the typical shell content in sea-sand has little influence on the 28 d compressive strength of sea-sand concrete [34, 35]. In this experiment, the 28 d compressive strength of ordinary concrete is a little higher than SSC, which results in that the average bond strength of BFRP bars ordinary concrete is slightly higher than that of BFRP bars SSC.

3.3.3 Effects of bar type

Table 4 and Figure 6(c) show that there is little variation in bond strength of the steel rebar and BFRP bar. The bond strength of steel rebar specimen increases by 5.2% compared to the BFRP bar specimen.

This is mainly attributed to the difference of transverse ribs of reinforcements. The transverse rib of BFRP bar has smaller slope and smaller biting effect than the steel bars. As the shear strength and surface hardness of BFRP bars are lower than SSC, the cross ribs of bars will be worn to some extent during sliding, which reduces the interfacial mechanical biting force [11]. By contrast, the shear strength and surface hardness of the steel rebars are higher than that of SSC, and the cross-rib biting effect is significant and the mechanical biting force is larger. Moreover, in seawater sea-sand, the chloride ions cause slight corrosion expansion of the steel rebars, which is also beneficial to the improvement of bonding strength.

3.3.4 Effect of stirrups restraint

From Table 4 and Figures 6(d) and (e), it can be seen that the stirrups restraint may enhance the interfacial bond strength in a certain degree, and also the lifting amplitude increases with the increase of the number of stirrups. The bond strength was increased by 3.7% and 11.5% for 12 mm BFRP bar specimens with one stirrup restraint and two stirrups restraint, respectively, compared with the specimens without stirrups restraint. The bond strength was improved by 15.8% and 28.3% for 16 mm BFRP bar specimens with one stirrups restraint and two stirrups restraint, respectively, compared with the specimens without stirrups restraint. The above phenomena are mainly caused by the stirrup restraint of the SSC in core area [26, 27]. Stirrup restraint can change the failure mode of pull-out specimens to splitting-pulling failure from splitting failure or even pull-out failure. When the pull-out load is small, the small cracks will appear in the specimens, while in the core area the restraint effect of stirrup on the concrete can efficiently restrain the radial development of cracks. Stirrup will deform with the SSC, bearing the circumferential tensile stress induced by the wedge-shaped block in front of the rib of the reinforcement material, thus raising the splitting strength of the concrete. The stirrups stress increases dramatically again as the crack develops to the surface of the specimen, which restrains the longitudinal development of cracks, and finally improves the bonding strength of the interface.

4 Bond-slip constitutive model

Currently, the classical bond-slip constitutive models of FRP bar and concrete mainly include BPE model [14], MBPE model [15], CMR model [16], continuous curve model [17] and GFRP/steel bars strand composite bar bond-slip model [18]. From the analysis of bond-slip curves in Section 3.2, it can be seen that some existing constitutive models have some shortcomings compared with real τ-s curves. The BPE model [14] was originally put forward based on deformed reinforcement. It has some limitations when applied to FRP bars. There is a horizontal section in the curve which is inconsistent with the actual situation and the calculation results are quite discrete. The MBPE model [15] took into account the impact of surface form of FRP bars to the bond strength and ignores the horizontal section of the curve, but without considering the effect of the bar diameter. The CMR model [16] was based on the performance of the structure at the service stage and defined only the rising segment of the curve precisely. It did not accurately describe the micro slip stage and the bond stress in the residual stage was simplified to a constant value which is inconsistent with the reality, although the physical concept of continuous curve model [17] was clear. GFRP/ steel wire composite rebar bond-slip model [18] was improved on the basis of the above model, where the residual stage was well defined, but the descent stage was simplified to linear without excessive smoothing.

4.1 Model establishment

According to the experimental results and references [14, 17, 18], the bond-slip constitutive model of ribbed BFRP bars SSC is obtained and illustrated in Figure 7. OA segment curves rise linearly and is defined as micro slip stage; AB segment curves rise non-linearly to peak bond strength, defined as slip stage, based on rising segment of reference [14] model; BC segment curves descend non-linearly to the lowest point, defined as falling segment of reference [17] model; CD segment curves take the form of reciprocating cycle attenuation of sinusoidal function, defined as residual stage, reference [18] model residual stage. Model of bond slip for ribbed BFRP bars SSC is shown in Eq. (2):

Micro slip stage:

                    (2a)

Slip stage:

  (2b)

Descent stage:

,       (2c)

Residual stage:

,                 (2d)

where τ₁, τ_u, τ₃ are the bond stresses at points A, B and C in Figure 7; s₁, s_u, s₃ are the corresponding slip amounts in Figure 7; τ₁, τ_u, τ₃ and α, A, B, C, D are the parameters obtained from the bond-slip test results.

4.2 Model solution

The parameters can be obtained by substituting the test values of A, B and C points of the bond-slip test curve of ribbed and stiffened specimens for the pull-out failure into Eq. (2). The statistics of test characteristic values at curve boundary points and fitting values of parameters in the constitutive model are shown in Table 5. Based on the fitting values of model parameters in Figure 7 and Table 5,the conclusion can be drawn that the bond-slip constitutive model of ribbed BFRP bars SSC in this test is as follows:

Micro slip stage:

                     (3a)

Slip stage:

  (3b)

Descent stage:

,       (3c)

Residual stage:

,             (3d)

where the meanings of each parameter are the same as that in Eq. (2).

Figure 7 Bond-slip constitutive model of ribbed BFRP bars and SSC

4.3 Model validation

By substituting the relevant data in Table 5 into Eq. (2), the fitting bond-slip curves of each specimen can be obtained. The comparison between the test curve of bond slip of each test piece and the fitting model is shown in Figure 8. It is observed that the fitting model can better reflect the smooth transition characteristics and non-linear trend of the downward section of the test curve. It is that the bond stress decreases gradually for a short time and then decreases sharply. In addition, the cyclic reciprocating attenuation phenomenon of bond strength in the residual section is also well simulated. Therefore, the model obtained in this paper is better matched with the test curves, which to some extent reflects the characteristics of bond-slip curves and the complete stress process of BFRP bars. Due to the limited number of available specimens, it is necessary to design bond specimens with different factors to further study each parameter to further optimize the bond-slip model.

5 Conclusions

In this study, the bonding property of BFRP bar and SSC is investigated by pull-out test, and these influence factors, the types and diameters of FRP bars, the types of concrete and stirrups restraint. Based on the research results, the main conclusions are as follows:

1) The failure modes of the specimens include splitting failure, splitting-pull out failure and pull-out failure. Pull-out failures are primarily focused on the specimens with smaller diameter of the bars. The splitting failure is mainly concentrated on specimens with larger diameter of the bars, while splitting-pull out failure is primarily focused on specimens with larger diameter restrained by stirrups. The restraint of the stirrup can improve the bond strength of the interface to some extent, and the enhancement increases with the increase of the number of stirrups. Decreasing the diameter of the bars is beneficial to improve the bonding performance of BFRP bars SSC.

Table 5 Eigenvalues and parameter fitting values of test data

Figure 8 Comparison of bond-slip test curve with fitting model:

2) The bond-slip curve of BFRP ribbed bars includes micro slip stage, slip stage, descent stage and residual stage. The residual stage bond strength shows the cycle of sine wave attenuation. Different from FRP bars, when the bond-slip curve of the steel rebar specimens enters the falling stage, the bond stress gradually decreases to the minimum value, and there is no residual stage.

3) The bond strength of BFRP bars is slightly lower than that of the steel specimens, which is mainly due to the different shear capacity and slope of the two cross ribs. Replacement of freshwater river sand by seawater sand has no obvious impact on the bonding performance between BFRP bars and concrete in the short term.

4) The same constitutive model as that of ordinary concrete can be used for the micro slip stage and slip stage of bond-slip curve of BFRP bars SSC. The bond-slip constitutive model from the test results and relevant literatures has good consistency with the measured curves. Due to the limited number of available specimens, further experiments are needed in the future to investigate the parameters in the constitutive model in depth.

Contributors

The overarching research goals were developed by SU Xun, YIN Shi-ping, ZHAO Ying-de, and HUA Yun-tao. SU Xun and YIN Shi-ping provided the measured landslides displacement data, and analyzed the measured data. SU Xun, YIN Shi-ping, and ZHAO Ying-de established the models and calculated the predicted displacement. SU Xun and HUA Yun-tao analyzed the calculated results. The initial draft of the manuscript was written by SU Xun, YIN Shi-ping, ZHAO Ying-de, and HUA Yun-tao. All authors replied to reviewers’ comments and revised the final version.

Conflict of interest

SU Xun, YIN Shi-ping, ZHAO Ying-de, and HUA Yun-tao declare that they have no conflict of interest.

References

[1] MOHAMMED T U, HAMADA H, YAMAFJI T. Performance of seawater-mixed concrete in the tidal environment [J]. Cement and Concrete Research, 2004, 34(4): 593-601. DOI: 10.1016/j.cemconres.2003.09.020.

[2] XIAO J, QIANG C, NANNI A. Use of sea sand and seawater in concrete construction: Current status and future opportunities [J]. Construction and Building Materials, 2017, 155(11): 1101-1111. DOI: 10.1016/j.conbuildmat.2017.08. 130.

[3] ROBERT M, BENMOKRANE B. Combined effects of saline solution and moist concrete on long-term durability of GFRP reinforcing bars [J]. Construction and Building Materials, 2013, 38(1): 274-284. DOI: 10.1016/j.conbuildmat. 2012.08.021.

[4] HASANZADE-INALLU A, ZARFAM P, NIKOO M. Modified imperialist competitive algorithm-based neural network to determine shear strength of concrete beams reinforced with FRP [J]. Journal of Central South University, 2019, 26(11): 3156-3174. DOI: https://doi.org/10.1007/ s11771-019-4243-z.

[5] SOONG W H, RAGHAVAN J, RIZKALLA S H. Fundamental mechanisms of bonding of glass fiber reinforced polymer reinforcement to concrete [J]. Construction and Building Materials, 2011, 25(6): 2813-2821. DOI: 10.1016/j.conbuildmat.2010.12.054.

[6] ARIAS J, VAZQUEZ A, ESCOBAR M M. Use of sand coating to improve bonding between GFRP bars and concrete [J]. Journal of Composite Materials, 2012, 46(18): 2271-2278. DOI: 10.1177/0021998311431994.

[7] YANG Qi-fei, LU Xiao-hua, XIONG Guang-jing. Double pull specimen more suitable for measuring bond-slip relationship of FRP-to-concrete interface [J]. Journal of Central South University of Technology, 2010, 17(2): 400-405. DOI: 10.1007/s11771-010-0059-6.

[8] REFAI A E, AMMAR M A, MASMOUDI R. Bond performance of basalt fiber-reinforced polymer bars to concrete [J]. Journal of Composites for Construction, 2015, 19(3): 04014050.1. DOI: 10.1061/(ASCE)CC.1943-5614. 0000487.

[9] ROLLAND A, QUIERTANT M, KHADOUR A. Experimental investigations on the bond behavior between concrete and FRP reinforcing bars [J]. Construction and Building Materials, 2018, 173(1): 136-148. DOI: 10.1016/j.conbuildmat.2018.03.169.

[10] GU Xing-yu, YU Bin, WU Ming. Experimental study of the bond performance and mechanical response of GFRP reinforced concrete [J]. Construction and Building Materials, 2016, 114(1): 407-415. DOI: 10.1016/j.conbuildmat.2016. 03.211.

[11] MAZAHERIPOUR H, BARROS J, SENA-CRUZ J M. Experimental study on bond performance of GFRP bars in self-compacting steel fiber reinforced concrete [J]. Composite Structures, 2013, 95: 202-212. DOI: 10.1016/j.compstruct. 2012.07.009.

[12] ALTALMAS A, REFAI A E, ABED F. Bond degradation of basalt fiber-reinforced polymer (BFRP) bars exposed to accelerated aging conditions [J]. Construction and Building Materials, 2015, 81: 162-171. DOI: 10.1016/ j.conbuildmat.2015.02.036.

[13] REN Zhen-hua, LIU Han-long, ZHOU Feng-jun. Double linear strain distribution assumption of RC beam strengthened with external-bonded or near-surface mounted fiber reinforced plastic [J]. Journal of Central South University, 2012, 19(12): 3582-3594. DOI: 10.1007/s11771-012-1446-y.

[14] ELIGEHAUSEN R, POPOV E P, BERTERO V V. Local bond stress-slip relationships of deformed bars under generalized excitations [R]. Calif: College of Engineering, University of California, 1982.

[15] COSENZA E, MANFREDI G, REALFONZO R. Analytical modelling of bond between FRP reinforcing bars and concrete [C]// TAERWE L. Proc 2nd Int RILEM Symp (FRPRCS-2), 1995: 164-171.

[16] COSENZA E, MANFREDI G, REALFONZO R. Behavior and Modeling of Bond of FRP Rebars to Concrete [J]. Journal of Composites for Construction, 1997, 1(2): 40-51. DOI: 10.1061/(ASCE)1090-0268(1997)1:2(40).

[17] GAO Dan-ying. The constitutive models for bond slip relation between FRP rebars and concrete [J]. Industrial Construction, 2003, 33(7): 41-43. (in Chinese)

[18] HAO Qing-duo, WANG Yan-lei, HOU Ji-lin. Bond-slip constitutive model between GFRP/steel wire composite rebars and concrete [J]. Engineering Mechanics, 2009, 26(5): 62-72. (in Chinese)

[19] DONG Zhi-qiang, WU Gang, ZHU Hong. Mechanical properties of seawater sea-sand concrete reinforced with discrete BFRP-Needles [J]. Construction and Building Materials, 2019, 206(5): 432-441. DOI: 10.1016/ j.conbuildmat.2019.02.029.

[20] LI Li-juan, HOU Bin, LU Zhong-yu. Fatigue behavior of sea sand concrete beams reinforced with basalt fiber-reinforced polymer bars [J]. Construction and Building Materials, 2018, 179(8): 160-171. DOI: https://doi.org/10.1016/j.conbuildmat. 2018.05.218.

[21] LI Y L, TENG J G, ZHAO X L. Theoretical model for seawater and sea sand concrete-filled circular FRP tubular stub columns under axial compression [J]. Engineering Structures, 2018, 160: 71-84. DOI: 10.1016/j.engstruct.2018. 01.017.

[22] ZHANG Qing-tian, XIAO Jian-zhuang, LIAO Qing-xiang. Structural behavior of seawater sea-sand concrete shear wall reinforced with GFRP bars [J]. Engineering Structures, 2019, 189(1): 458-470. DOI: 10.1016/j.engstruct.2019.03.101.

[23] DONG Zhi-qiang, WU Gang, ZHAO Xiao-ling. Durability test on the flexural performance of seawater sea-sand concrete beams completely reinforced with FRP bars [J]. Construction and Building Materials, 2018, 19(12): 671-682. DOI: 10.1016/j.conbuildmat.2018.10.166.

[24] XIAO Jian-zhuang, LIAO Qing-xiang, ZHANG Qing-tian. Bond behavior between seawater sea-sand recycled aggregate concrete and glass-fiber-reinforced polymer bars [J]. Journal of Tongji University (Natural science), 2018, 46(7): 884-890, 971. (in Chinese)

[25] XU Jin-jin, YANG Shu-tong, LIU Zhi-ning. Study on the bond performance between CFRP bars and alkali-activated slag seawater and sea sand concrete [J]. Engineering Mechanics, 2019, 36(S1): 175-183. (in Chinese)

[26] DONG Zhi-qiang, WU Gang, ZHAO Xiao-ling. Long-term bond durability of fiber-reinforced polymer bars embedded in seawater sea-sand concrete under ocean environments [J]. Journal of Composites for Construction, 2018, 22(5): 04018042.1. DOI: 10.1061/(ASCE)CC.1943-5614.0000876.

[27] DONG Zhi-qiang, WU Gang, ZHAO Xiao-ling. Bond durability of steel-FRP composite bars embedded in seawater sea-sand concrete under constant bending and shearing stress [J]. Construction and Building Materials,2018, 192(12): 808-817. DOI: 10.1016/j.conbuildmat.2018.10.154.

[28] ASTM-11a. Standard test method for tensile properties of fiber-reinforced polymer matrix composite bars [S]. ASTM D7205, West Conshohocken, PA, 2011.

[29] ASTM D1141-98(13). Standard practice for the preparation of substitute ocean water [S]. West Conshohocken: ASTM International, 2008.

[30] ACI Committee 440.3R-12. Guide test methods for fiber-reinforced polymers (FRPs) for reinforcing or strengthening concrete structures [S]. American Concrete Institute, 2012.

[31] GB/T50152-2012. Standard for test methods of concrete structures [S]. China Academy of Building Science, 2012. (in Chinese)

[32] FIRAS S A, GILLES F, ROBERT L R. Bond between carbon bre-reinforced polymer (CFRP) bars and ultra high performance bre reinforced concrete (UHPFRC): Experimental study [J]. Construction and Building Materials, 2011, 25(2): 479-485. DOI: 10.1016/j.conbuildmat.2010. 02.006.

[33] YANG Shu-tong, YANG Chao, HUANG Mei-lin, LIU Yang, JIANG Ji-tong, FAN Guo-xi. Study on bond performance between FRP bars and seawater coral aggregate concrete [J]. Construction and Building Materials, 2018, 173(6): 272-288. DOI: https://doi.org/10.1016/j.conbuildmat.2018.04.015.

[34] YANG E I, YI S T, LEEM Y M. Effect of oyster shell substituted for fine aggregate on concrete characteristics: Part I. Fundamental properties [J]. Cement and Concrete Research, 2005, 35(11): 2175-2182. DOI: 10.1016/j.cemconres.2005. 03.016.

[35] RICHARDSON A E, FULLER T. Sea shells used as partial aggregate replacement in concrete [J]. Structural Survey, 2013, 31(5): 347-354. DOI: 10.1108/SS-12-2012-0041.

(Edited by YANG Hua)

中文导读

BFRP筋与海水海砂混凝土黏结性能的试验研究

摘要：纤维增强聚合物材料(FRP)与海水海砂混凝土(SSC)结合可以解决海洋工程建设中的河砂短缺问题。本文通过拉拔试验研究了BFRP筋与海水海砂混凝土的粘结性能。考虑了筋材类型、筋材直径、混凝土类型和箍筋约束等参数的影响。减小筋材直径有利于提高粘结性能，海水海砂的利用对混凝土的粘结性能影响不大，BFRP的粘结强度比钢筋略低，但差别较小。箍筋约束可以改变试件的破坏模式并在一定程度上提高界面粘结应力。带肋BFRP筋的粘结-滑移曲线包括微滑阶段、滑移阶段、下降阶段和残余阶段，且残余阶段粘结应力呈正弦式循环衰减形式。此外，基于实验结果和相关文献得到了BFRP筋和海水海砂混凝土的粘结滑移模型，预测的曲线与实测曲线有很好的一致性。

关键词：玄武岩纤维增强聚合物(BFRP)；海水海砂混凝土(SSC)；粘结-滑动曲线；本构模型

Foundation item: Project(BE2019642) supported by the Jiangsu Provincial Key Research and Development Program, China

Received date: 2020-07-20; Accepted date: 2021-05-05

Corresponding author: YIN Shi-ping, PhD, Professor; Tel: +86-15262013916; E-mail: yinshiping2821@163.com; ORCID: https://orcid. org/0000-0001-8304-5914